Infant mortality, the death of baby under 1 year, is an important indicator of child health. In the UK, infant mortality had been decreasing continuously over 100 years until 2015. From 2015 to 2017, a gradual increase has been observed and was hypothesised to be due to increased deprivation. It is known that babies born to socially/economically deprived parents (or parents living in deprived areas) have a higher likelihood of dying. Previous reports also found that the difference in infant mortality due to deprivation has not been narrowed from 1950 to 2018. However, it is important to see whether the trend is robust after considering other potential explanations. It is also important to understand how deprivation may influence infant mortality. Some hypotheses are parental age at delivery, maternal illnesses, smoking and drinking during pregnancy, and pregnancy complications. This study will verify any socioeconomic disparity in infant mortality and the whether such disparity can be explained by maternal and pregnancy factors. While upstreaming factors for health disparity (such as deprivation) is harder and takes a longer time to improve, maternal and pregnancy factors can be changed through healthcare services in the exisiting framework. This study will contribute to the reduction of socioeconomic disparity in infant mortality through modifying the intermediate factors.
This study aims to study whether the socioeconomic disparity in infant mortality (under 1 year) has widened, narrowed, or remained stable and whether this can be explained by potential mediators: maternal age and ethnicity, smoking and drinking during pregnancy, maternal illnesses, and pregnancy complications using the CPRD Gold data. A cohort of pregnant women in the CPRD pregnancy register will be established. Stillbirths will be excluded. Deaths of their children will be ascertained through the linkage to the ONS death registration. Area-based socioeconomic status (relative deprivation) will be assessed based on the linked small-area data (twentiles of English Index of Multiple Deprivation). Child sex, year of birth, maternal age at delivery, any pregnancy complications and birth characteristics (gestational diabetes, pregnancy hypertension, (pre-)eclampsia, preterm labour, mode of delivery, and maternal prior history of mental disorders (depression and anxiety), disability, and cancer will be extracted from the primary care data. Poisson regression with robust standard errors will be used to study the trend of socioeconomic disparity. Infant all-cause death will be the outcome variable. Child sex, year of delivery, area-based deprivation, and the interaction between year of delivery and deprivation will be the predictors. Interaction tests will be used to indicate any changes in disparity. The same model will be replicated with potential mediators as additional covariates. The percentage decrease in the effect size of the interaction terms indicates the role of the mediators. As sensitivity analysis, the above primary analysis will be replicated with: 1. Generalised additive model for potential nonlinear trend; 3. Neonatal (0-27 days) and post-neonatal (28-365 days) death as outcomes; and 4. Mothers with mental health disorders, disabilities, and cancers prior to delivery excluded.
All-cause infant (<1 year old) mortality will be primary outcome; neonatal (0-27 days), and post-neonatal (28-365 days) will be secondary outcomes used in sensitivity analyses.
Frederick Ho - Chief Investigator - University of Glasgow
Frederick Ho - Corresponding Applicant - University of Glasgow
Bhautesh Jani - Collaborator - University of Glasgow
Daniel Mackay - Collaborator - University of Glasgow
Jill Pell - Collaborator - University of Glasgow
Kenneth Man - Collaborator - University College London ( UCL )
Max Allan - Collaborator - University of Glasgow
Paul Welsh - Collaborator - University of Glasgow
CPRD Mother-Baby Link;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Pregnancy Register